Hacker News new | ask | show | jobs
by TaylorAlexander 2079 days ago
They sell printers which “solve” that problem by locking you in to one type of filament.

I would love to see automatic print failure detection (with a camera, even a micro camera at the nozzle).

Figuring out the correct settings for unknown plastic seems like it would be hard to solve, but perhaps there’s some base material properties that could be determined automatically which might help. I’ve noticed that the Prusa 3D printer software has a sizable database of filament brands in its material settings section. I’d be curious how foolproof those settings all are for each brand of filament. But of course a manufacturer could change formulation at any time...

I solve this problem by only buying one of two brands and keeping custom settings for them.

I started my career in CNC machining. People always thought that meant software did everything for me, but no I had to choose tool sizes, spindle RPM, feed rates, depth of cut, etc. The CNC is not a magic machine that makes parts for you, it’s a tool. So I guess I’m a bit biased for thinking that current 3D printers are super useful tools that require a certain amount of operator competency. You cannot put any random filament in and expect good results, and actually that might be a hard problem to solve on the back end. Much easier to pay attention to what brand you order and feed in good quality stuff.

But idk. You’re right that they’re fiddly AF and I’m just used to it. If they could solve those problems it would surely help adoption.

2 comments

>I would love to see automatic print failure detection (with a camera, even a micro camera at the nozzle).

You probably already know about it, but Spaghetti Detective is pretty decent at detecting print failures - although as you say it'd be great to have this built in (and more reliable...)

Hey I didn’t know about that! Seems interesting. Thanks!
This feels like the filament should come with a barcode that tells the printer all the requirements
There’s some printer-specific factors that affect ideal temperature settings. Hot end geometry varies by printer and can affect ideal print temperature. Also they used to often have inaccurate temperature sensing, though that is fixable and perhaps things have improved there. I don’t know how much these factors would affect things, but if the variance across products is too great, filament manufacturers couldn’t solve the problem with pre defined settings.

Other factors that affect print quality are: filament diameter consistency (must be very precise), absolute filament diameter (must be known), nozzle wear (changes over time), environmental humidity and air temperature, and air flow. Michigan, Florida, and Arizona might require different settings. And then part geometry matters a lot. Someone up the comment thread mentioned curling. This can happen even with perfect temperature settings if part geometry is not ideal. Once I was totally unable to print a large rectangular part without curling until I replaced the 0.4mm nozzle with a 0.8mm nozzle. The thicker layers eliminated layer delamination that allowed the curling.

It really takes a bit of learning and experimentation to get things working sometimes. It’s not a perfect process by any means.